Predictor

main

nlper.predictor.__init__.main(text: str)

Executes the text prediction pipeline, obtaining the summarization

Parameters

text (str) – Text to summarize

application

class nlper.predictor.application.Application(text: str, config_path: str = 'resources/model_files/predict_config.yaml')

Text predict application which obtains text summarization. Starts by initializing objects with cleaning utils, json reader and vocabulary config.

Parameters

• text (str) – Text to summarize

• config_path (str) – Path to yaml config file, by default loads example config

predict() → None

Calls model predict method and creates attention heatmap.

prepare_model() → None

Initializes and loads saved model for prediction.

• If config file specifies use_dummy_model as True, the model is initialized with random weights.

prepare_text() → None

Prepares text for prediction.

prepare_vocab() → None

Prepares vocabulary by loading it from vocab_path specified in yaml config file.

• Assigns parameters called itos and stoi of VocabConfig class.

• Sets text size to numbers of words in vocabulary

run() → None

Executes text prediction process.